The Case for Power Management in Web Servers

Bohrer, Pat; Elnozahy, E. N. Mootaz; Keller, Teresa; Kistler, Michael; Lefurgy, Charles; McDowell, C.; Rajamony, Ram

doi:10.1007/978-1-4757-6217-4_14

Cited by 247 publications

(193 citation statements)

References 13 publications

Supporting

Mentioning

189

Contrasting

Order By: Relevance

“…This leads to sub-optimal resource utilization. Bohrer et al have studied real webserver workloads from sports, e-commerce, financial, and internet proxy clusters to find that average server utilization varies between 11% and 50% [3]. Such inefficient provisioning leads to relatively large hardware and operations costs when compared to the actual workload handled by the data center.…”

Section: Introductionmentioning

confidence: 99%

pMapper: Power and Migration Cost Aware Application Placement in Virtualized Systems

Verma¹,

Ahuja²,

Neogi³

2008

Lecture Notes in Computer Science

527

437

View full text Add to dashboard Cite

Abstract. Workload placement on servers has been traditionally driven by mainly performance objectives. In this work, we investigate the design, implementation, and evaluation of a power-aware application placement controller in the context of an environment with heterogeneous virtualized server clusters. The placement component of the application management middleware takes into account the power and migration costs in addition to the performance benefit while placing the application containers on the physical servers. The contribution of this work is two-fold: first, we present multiple ways to capture the cost-aware application placement problem that may be applied to various settings. For each formulation, we provide details on the kind of information required to solve the problems, the model assumptions, and the practicality of the assumptions on real servers. In the second part of our study, we present the pMapper architecture and placement algorithms to solve one practical formulation of the problem: minimizing power subject to a fixed performance requirement. We present comprehensive theoretical and experimental evidence to establish the efficacy of pMapper.

show abstract

Section: Introductionmentioning

confidence: 99%

pMapper: Power and Migration Cost Aware Application Placement in Virtualized Systems

Verma¹,

Ahuja²,

Neogi³

2008

Lecture Notes in Computer Science

527

437

View full text Add to dashboard Cite

show abstract

“…Therefore, due to the sleep power and frequency-independent active power, there is an energy efficient processing frequency at which the energy consumption to process a request is minimized [28]. Since the overhead of turning on/off a server is large [1], we assume in this paper that the deployed servers are always on and the sleep power P s is not manageable (i.e., always consumed). Thus, the energy efficient frequency can be easily found as:…”

Section: Power Modelmentioning

confidence: 99%

“…However, the energy management for servers in data centers, where heat generated and cooling costs are big problems, have caught people's attention recently. In [1], Bohrer et al presented a case of managing power consumption in web servers. Elnozahy et al evaluated a few policies that combine dynamic voltage scaling (DVS) [24,25] on individual server and turning on/off servers for cluster-wide power management in server farms [5,14].…”

Section: Introductionmentioning

confidence: 99%

Energy Efficient Configuration for QoS in Reliable Parallel Servers

Zhu

Melhem

Mossé

2005

Dependable Computing - EDCC 5

View full text Add to dashboard Cite

Abstract. Redundancy is the traditional technique used to increase system reliability. With modern technology, in addition to being used as temporal redundancy, slack time can also be used by energy management schemes to scale down system processing speed and supply voltage to save energy. In this paper, we consider a system that consists of multiple servers for providing reliable service. Assuming that servers have self-detection mechanisms to detect faults, we first propose an efficient parallel recovery scheme that processes service requests in parallel to increase the number of faults that can be tolerated and thus the system reliability. Then, for a given request arrival rate, we explore the optimal number of active severs needed for minimizing system energy consumption while achieving k-fault tolerance or for maximizing the number of faults to be tolerated with limited energy budget. Analytical results are presented to show the trade-off between the energy savings and the number of faults being tolerated.

show abstract

“…Energy consumption is emerging as a critical issue in the design of computing systems [5,14,20,21,28,32,41]. The goals of energy-aware system design include saving energy without sacrificing performance, and supporting flexible, dynamic trade-offs between energy consumption and performance.…”

Section: Introductionmentioning

confidence: 99%

On the energy consumption and performance of systems software

Grosu

Sehgal

et al. 2011

Proceedings of the 4th Annual International Conference on Systems and Storage

View full text Add to dashboard Cite

Models of energy consumption and performance are necessary to understand and identify system behavior, prior to designing advanced controls that can balance out performance and energy use. This paper considers the energy consumption and performance of servers running a relatively simple file-compression workload. We found that standard techniques for system identification do not produce acceptable models of energy consumption and performance, due to the intricate interplay between the discrete nature of software and the continuous nature of energy and performance. This motivated us to perform a detailed empirical study of the energy consumption and performance of this system with varying compression algorithms and compression levels, file types, persistent storage media, CPU DVFS levels, and disk I/O schedulers. Our results identify and illustrate factors that complicate the system's energy consumption and performance, including nonlinearity, instability, and multi-dimensionality. Our results provide a basis for future work on modeling energy consumption and performance to support principled design of controllable energy-aware systems.

show abstract

The Case for Power Management in Web Servers

Cited by 247 publications

References 13 publications

pMapper: Power and Migration Cost Aware Application Placement in Virtualized Systems

pMapper: Power and Migration Cost Aware Application Placement in Virtualized Systems

Energy Efficient Configuration for QoS in Reliable Parallel Servers

On the energy consumption and performance of systems software

Contact Info

Product

Resources

About